Gemini Chrome FAQ
Common questions about the Meridian ↔ Gemini bridge.
Why does the bridge add latency?
Every request passes through Meridian's proxy layer for tokenization, context trimming, and response streaming. This adds 80–200ms on top of Gemini's native inference time. We prioritize correctness over raw speed — your session state and rate limits stay consistent.
What cookies does the bridge set?
A single HttpOnly session cookie (__Host-sid) with SameSite=Strict. No third-party cookies, no tracking pixels, no fingerprinting. The cookie expires when you close the tab.
Is there a cost to use the bridge?
Meridian Pro subscribers ($12/mo) get unlimited bridge access. Free tier users receive 50 bridge requests per day. Each request consumes one quota slot regardless of token count. Quota resets at midnight UTC.
Why use the bridge instead of the native API?
The bridge gives you unified billing, Meridian's context window management, automatic fallback routing, and audit logging. If you call Gemini directly you lose session continuity and bypass our safety guardrails. For teams, the bridge enforces organization-wide rate limits that raw API keys cannot.
When should I fall back to native gemini-3-flash?
Fall back when you need sub-50ms latency, when you're sending raw binary payloads the bridge doesn't parse, or when Meridian is in maintenance mode. Set the header X-Meridian-Bypass: 1 and the request will route directly to Gemini without proxy processing.